Long range surface plasmon resonance for increased sensitivity in living cell biosensing through greater probing depth

Long range surface plasmon resonance for increased sensitivity in living cell biosensing through greater probing depth

Sensors and Actuators B 174 (2012) 94–101 Contents lists available at SciVerse ScienceDirect Sensors and Actuators B: Chemical journal homepage: www...

690KB Sizes 0 Downloads 35 Views

Sensors and Actuators B 174 (2012) 94–101

Contents lists available at SciVerse ScienceDirect

Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb

Long range surface plasmon resonance for increased sensitivity in living cell biosensing through greater probing depth Vincent Chabot a , Yannick Miron b , Michel Grandbois b , Paul G. Charette a,∗ a b

Département de génie électrique et génie informatique, Faculté de génie, Université de Sherbrooke, 2500 Boulevard Université, Sherbrooke, Québec, J1K 2R1, Canada Département de Pharmacologie, Faculté de Médecine, Université de Sherbrooke, 3001 12e Avenue Nord, Sherbrooke, Québec, J1H 5N4, Canada

a r t i c l e

i n f o

Article history: Received 29 June 2012 Received in revised form 16 August 2012 Accepted 20 August 2012 Available online 25 August 2012 Keywords: Surface plasmon Cell biosensor Long range plasmon Evanescent field scattering

a b s t r a c t Long range surface plasmon resonance (LRSPR) was used to improve over standard SPR for the detection and monitoring of toxicity with living cells. With a greater evanescent field penetration depth into the sensing medium and increased sensitivity to optical scattering, LRSPR shows a higher sensitivity for measurements involving molecular content redistribution associated with cellular morphology changes. LRSPR and SPR sensitivity to real and imaginary refractive index component variations was evaluated separately using calibrated oil solutions and dilutions of polystyrene beads, respectively. An increase in reflectance at the minimum along with a broadening of the curve was observed for LRSPR as a result of evanescent field scattering. Angular scans for SPR were marginally affected, due to an evanescent field too shallow for the beads to produce significant scattering. Methylene blue dilutions in water were used to evaluate sensitivity to complex refractive index variations. Finally, LRSPR and SPR were used to measure the response of a HEK-293 cell monolayer under stimulation at different concentrations by a toxin, lipopolysaccharides. The experimental results presented confirm that living cells must be modeled as a dielectric medium with a complex refractive index in LRSPR measurements and that LRSPR exhibits a 50% greater sensitivity compared to standard SPR for toxicity measurements based on living cells. © 2012 Elsevier B.V. All rights reserved.

1. Introduction There is a wide range of applications for toxicity biosensors, notably for environmental studies, food quality control and medical applications. To be suitable for toxicity detection, a sensor should demonstrate a fast response and a high sensitivity across a broad spectrum of compounds. Since living cells respond readily to a range of changes in their immediate environment, they can be used as the sensing element at the heart of a toxicity sensor. When combined with surface plasmon resonance (SPR), a wide variety of compounds can be detected rapidly at very low concentrations [1]. Surface plasmons are charge density oscillations at the interface between a metal and a dielectric. When transverse magnetic (TM) polarized light with the appropriate wave vector is reflected off the back side of a thin metal film, for example using a prism in the well-known Kretschmann configuration, the light couples resonantly to the surface plasmons on the opposite side of the metal [2]. This resonance, which appears as a sharp dip in reflectance as a function of incidence angle and interrogation

∗ Corresponding author. Tel.: +1 819 821 8000x63861; fax: +1 819 821 7937. E-mail address: [email protected] (P.G. Charette). 0925-4005/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.snb.2012.08.028

wavelength, generates an evanescent field in the dielectric medium above the metal film. When the refractive index of the dielectric changes within the range of the evanescent field, the resonance coupling conditions (incident angle/wavelength combination) change accordingly. In the most common configuration, the wavelength is held constant while the incidence angle is adjusted to track the changing coupling conditions. This resonant relationship between reflectance and refractive index leads to SPR’s high sensitivity to small index changes in the vicinity of the metal surface. In most applications, one considers that the sensing dielectric medium has a refractive index that is purely real or, more rigorously, that only the real part of the possibly complex refractive index undergoes change during an experiment. In this case, the shape of the reflectance function and its minimum value attained at resonance are fairly constant over a wide range of coupling conditions. As a result, the reflectance function plotted against incidence angle at constant excitation wavelength will undergo a simple translation in response to a refractive index change in the dielectric. The sensor output is either the shift in the angular position of the reflectance minimum or the change in reflectance at a fixed angle. The widely accepted definition for SPR instrumentation bulk sensitivity at a particular wavelength is the maximum rate of

V. Chabot et al. / Sensors and Actuators B 174 (2012) 94–101

change of reflectance, R, with respect to bulk change in refractive index, nbulk , across all possible angles of incidence,  [3]:



Sbulk = max|

ıR ınbulk



(1)

As described above, sensitivity can be determined experimentally from the two sets of reflectance measurements acquired over the full range of incidence angles before and after a known bulk refractive index change: the maximum point-wise reflectance difference between the two sets divided by the index change yields the sensitivity. Most SPR studies use gold as the metallic layer, mainly due to its ease of functionalization and stability [4]. Moreover, gold is biocompatible and allows cellular monolayer growth when coated with appropriate adhesion promoters. Silver shows a more sensitive SPR response under certain conditions, but has low chemical stability in biological solutions [5] and is cytotoxic [6]. A common technique to combine the increased sensitivity of silver with the chemical stability of gold is to use a bimetallic configuration, where a fine layer of gold (∼5 nm) is deposited on top of a thicker silver layer [7]. This technique is employed in the work presented here. In practice, all dielectric materials absorb and scatter light to some degree and their full refractive index is complex, where the imaginary component of the index quantifies the attenuation in the medium [8]. Surface plasmons generate a non-uniform plane wave that travels parallel to the metal/dielectric interface in the direction of plasmon propagation with exponentially decaying amplitude in the direction normal to the interface. This evanescent wave can be scattered and/or absorbed like any other electromagnetic wave [9]. As a result, when the imaginary component of the dielectric refractive index is non-negligible (“lossy” material), the reflectance function plotted against incidence angle will undergo more than a simple shape-invariant translation in response to a change in the attenuation of the sensing medium [10]. In standard SPR, the optimal metal thickness is set to exactly balance the internal damping losses in the metal (and in the dielectric if the material is lossy) with the losses due to radiation back into the prism, resulting in near-zero reflectance at resonance [11]. However, in the case of a lossy dielectric medium, this balance is upset when the absorption in the dielectric changes, leading to nonzero reflectance values at resonance and a broadening of the SPR curve [10]. By taking advantage of this additional information, SPR sensing performance can be improved, as explained at the end of this section for the case of living cells. With a cellular monolayer grown directly on an SPR sensing substrate, it is possible to detect variations in the cellular environment by tracking changes in the average refractive index of the monolayer caused by changes in cellular morphology [12,13]. Indeed, cellular responses such as contraction, spreading, cell rounding and membrane blebbing translate into a change in the ratio of cell/medium volume within the evanescent field [14]. This in turn results in a change of the average refractive index detected by SPR and has been used in a wide range of cellular applications [15–19]. The evanescent field penetration depth into the dielectric medium is defined as the distance over which the amplitude of the evanescent electric field decreases to 1/e of its initial value, as shown in Fig. 1c. The penetration depth gives a good indication of the usable distance from the surface at which an SPR system is sensitive to changes in the dielectric (∼200 nm at optical frequencies). Cell height, however, is in the order of several microns or more. As a result, most of the cell volume is outside the sensing range in the standard SPR configuration. In order to improve the sensitivity of SPR by probing deeper into the cell body, we propose to use “long-range” surface plasmon resonance (LRSPR). In LRSPR, the plasmon modes on

95

opposing surfaces of the metal film are coupled together via a symmetric substrate configuration (dielectric–metal–dielectric or metal–dielectric–metal), as shown in Fig. 1a. When the two plasmons are coupled symmetrically with respect to their magnetic fields, the effective index of the surface plasmon mode decreases [2] and as a result, the difference between the refractive index of the surface plasmon mode and the dielectric also decreases, causing the plasmon mode to be less confined in the metal layer [20], thereby extending the penetration depth of the evanescent field into the dielectric medium above (Fig. 1c). Hence, compared to “standard” SPR, LRSPR shows a greater sensitivity to bulk refractive index variations in the dielectric medium [21] and has been used to detect thick analytes such as bacteria [22,23]. Moreover, the LRSPR resonance curve plotted against incidence angle has a much sharper dip occurring at a lower coupling angle (Fig. 1d). Due to their micrometric dimensions and non-uniform composition, cells scatter light [24–26]. These scattering losses translate into a strong modulation of the imaginary component of the sensing medium refractive index. As a result of the deeper penetration of the evanescent field, attenuation by the cell monolayer will significantly influence the reflectance response in LRSPR. In this manuscript, we demonstrate that the cellular response to an external stimulation observed with LRSPR must be modeled as a complex variation of the refractive index. We also show that LRSPR exhibits an increased overall sensitivity to cellular responses when compared with standard SPR. 2. Materials and methods 2.1. Numerical simulations Coupled Fresnel equations were used to model the reflectance response from substrates with the following film thicknesses and associated material parameters at 635 nm obtained from the literature [27,28]: 3 nm Ni (nNi = 2.000 + i3.774), 32 nm Ag (nAg = 0.136 + i4.002) and 5 nm Au (nAu = 0.179 + i3.067) for standard SPR substrates; 650 nm Teflon (nTeflon = 1.308), 32 nm Ag and 5 nm Au for LRSPR substrates. Differences between the modeled and experimental responses in the plots below can be attributed to small variations in film thickness and material parameters inherent to the fabrication process. 2.2. Substrate fabrication Standard glass microscope slides (Fisher Scientific, Canada) were used as base substrates. The glass slides were first cleaned in piranha solution (1 H2 O2 :3 H2 SO4 ) to remove any contaminants. For standard SPR substrates, a nickel adhesion layer (3 nm), silver layer (32 nm) and gold passivation layer (5 nm) were deposited subsequently without breaking vacuum between evaporation in an e-beam evaporator (Sloan Instruments, USA). Metal layer thickness was measured during deposition with a Sloan Omni III quartz microbalance. For LRSPR substrates, the cleaned slides were immersed in a 2% solution of perfluorodecyl trichlorosilane (Gelest, USA) dissolved in Opticlear (National Diagnostics, USA) for 5 min under nitrogen atmosphere then rinsed with isopropanol and dried with nitrogen. The substrates were then baked at 125 ◦ C for 20 min before spin-coating with Teflon AF1600 solution (Dupont, Canada). After the spin-coating process, the LRSPR substrates were dried for 1 h and baked at 160 ◦ C for 1 h. Metal deposition of 32 nm silver followed by 5 nm gold was performed at 150 ◦ C without breaking vacuum and the samples were subsequently brought down to room temperature over 5 h. Note that the Ni metal adhesion layer was not required for the LRSPR substrates and the substrates remained stable in saline solutions for several days. Both substrate

96

V. Chabot et al. / Sensors and Actuators B 174 (2012) 94–101

Fig. 1. (a) LRSPR substrate structure; (b) standard SPR substrate structure; (c) normalized (evanescent) electric field amplitude in the dielectric: circles represent the field “penetration depth”, where the electric field amplitude has decreased to 1/e of its initial value; (d) modeled reflectance as a function of angle for LRSPR and standard SPR substrates; (e) SPR experimental apparatus.

configurations are shown in Fig. 1a and b. Prior to cell culture, the substrates were coated to promote cellular adhesion by immersion in a solution of 5 ␮g/ml fibronectin (Sigma–Aldrich, USA) and 200 ␮g/ml gelatin (BD, USA) in water for 1 h at room temperature. 2.3. SPR apparatus A custom-built SPR apparatus was used for all experiments (Fig. 1e). The substrates were placed on a BK7 coupling prism (Melles Griot, USA) with refractive index matching fluid (Cargille Laboratories, USA). A Teflon fluid chamber was clamped on top of the SPR substrates. Light from a 4 mW stabilized laser diode centered at 635 nm (Thorlabs, USA) was reflected off the underside of the metal film, split by a polarizing beamsplitter (Edmund Optics,

USA) after reflection and measured with two biased photodetectors (Thorlabs, USA). Since the transverse electric (TE) polarization does not couple to the surface plasmons, it is used as a reference to remove signal dependence on laser power drift and other time-dependent noise sources. Angular scanning by simultaneous rotation of the laser and detectors was performed by rotational actuators (Thorlabs, USA). The system was controlled via a LabVIEW (National Instruments, USA) interface. Averaged data are shown as the mean value ± standard error of the mean (S.E.M.). 2.4. Reagents Oil solutions of calibrated refractive indices (Cargille Laboratories, USA) were used to measure sensitivity to purely real refractive

V. Chabot et al. / Sensors and Actuators B 174 (2012) 94–101

97

Fig. 2. Experimentally measured angular scans for oil solutions with refractive indices ranging from n = 1.33 to 1.355 by 0.005 increments for LRSPR (a) and SPR (b) substrates; (c) modeled and measured average (n = 3) sensitivities (slope) for both substrate types. Sensitivity is determined from the maximum point-wise reflectance difference between the two angular scans divided by the refractive index change (Eq. (1)). Standard error from the mean for SPR lies within the markers.

index variations. Solutions of polystyrene beads in 2.5% (w/v) concentration in water (Polysciences, USA) were used to measure sensitivity to purely imaginary refractive index variations. Dilutions of methylene blue (Fisher Scientific, Canada, absorption peak at 668 nm) in water were used to measure sensitivity to complex refractive index variations. Triton X-100 detergent was purchased from MP Biomedicals, Canada. Milli-Q water was used for all experiments. 2.5. Cell culture Human embryonic kidney-293 (HEK-293) cells (Qbiogene, USA) were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum (FBS), 2 mM l-glutamine, 2.5 ␮g/ml amphotericin B, 50IU/ml penicillin and 50 ␮g/ml streptomycin (Wisent, Canada) at 37 ◦ C in a 5% CO2 incubator. Prior to the experiments, the cells were plated on SPR and LRSPR surfaces in a 60 mm Petri dish and left to grow. The SPR substrates were observed through phase contrast microscopy to assess monolayer confluence and ensure reproducibility of the measurements. Once ∼90% confluence was reached (approx. 2 days), the cell-coated surfaces were mounted on the prism of the SPR setup for measurement. The cell culture medium was washed once and replaced with HEPES-Buffered Salt Solution (HBSS) consisting of 20 mM HEPES, 120 mM NaCl, 5.3 mM KCl, 0.8 mM MgSO4 , 1.8 mM CaCl2 and 11.1 mM dextrose, adjusted at pH 7.4.

2.6. Cell toxicity assay HEK-293 cellular response to lipopolysaccharides (LPS) was used as a toxicity model. LPS are involved in the inflammatory response induced by Gram-negative bacteria such as Escherichia coli and Salmonella and cause an important cellular response, often resulting in cellular death [29]. Studies have shown that this agent induces morphological changes such as cell rounding and membrane blebbing [30]. HEK-293 cells were stimulated with various concentrations of lipopolysaccharides (1, 5, 50 and 500 ␮g/ml) from E. coli (Sigma–Aldrich, USA). Phase contrast images of the LPStreated cells were taken using a Zeiss Axiovert 200 microscope with 20× magnification (Carl Zeiss, Canada). 3. Results To compare the sensitivity of LRSPR and SPR substrates to purely real refractive index variations, angular scans were performed for a range of oil solutions of calibrated refractive indices (n = 1.33–1.355, 0.005 increments). For LRSPR substrates (Fig. 2a), the angular position of the minimum undergoes a nearly pure translation in response to refractive index changes and a slight variation in the reflectance minimum value is observed caused by the increasing asymmetry of refractive indices between the dielectric (oil) and the Teflon layer. Similarly, for standard SPR substrates (Fig. 2b), a pure translation of the reflectance minimum in response to refractive index changes is observed. Fig. 2c shows the measured and

98

V. Chabot et al. / Sensors and Actuators B 174 (2012) 94–101

Fig. 3. Experimentally measured angular scans from LRSPR (a) and SPR (b) substrates for polystyrene bead solutions (1, 2 and 10 ␮m) and water; (c) modeled sensitivities (slope) for LRSPR and SPR substrates to variations in the imaginary component of the refractive index (0–0.035, in 0.001 increments).

modeled average maximal reflectance difference as a function of refractive index, referenced to the initial value (1.33). The sensitivity (maximal reflectance difference divided by the index change) is the slope of the curve. Bulk sensitivity is greater for LRSPR, as expected (dR/dn = 121 RIU−1 for LRSPR, 51 RIU−1 for SPR). To compare the sensitivity of LRSPR and SPR substrates to variations in the imaginary component alone of the refractive index (attenuation), solutions of polystyrene beads of different diameters (1, 2 and 10 ␮m) in water were used to induce particle size-dependent attenuation by scattering of the evanescent field [9]. This attenuation induces a strong modulation of the imaginary part of the refractive index, resulting in a variation in the minimum reflectance value and broadening of the curve for LRSPR (Fig. 3a). Interestingly, for standard SPR (Fig. 3b), the minimum reflectance value is relatively constant for the same experiments. This effect is mainly due to the nature of the experimental model (polystyrene beads in water), however, as the penetration depth of the evanescent field for standard SPR is too shallow for scattering by the beads to have a significant effect. Indeed, SPR is also sensitive to changes in the attenuation of the dielectric, as shown in the modeling results of Fig. 3c, though to a lesser degree. Since it is very difficult to estimate the exact attenuation at a particular bead diameter and concentration in this experimental model, actual measurements of sensitivity to variations in the imaginary component of the refractive index are not shown. The dynamic range of imaginary component values used in the modeling (Fig. 3c) covers small changes up to and including a complete loss of LRSPR response, a situation observed in practice in experiments with highly scattering media.

To compare the sensitivity of LRSPR and SPR substrates to complex refractive index variations, dilutions of Methylene Blue in water (0–10 mg/ml, 1 mg/ml increments) were used, as shown in Fig. 4. As expected, angular scans for 0, 5 and 10 mg/ml show that the different concentrations induce complex refractive index variations for LRSPR and SPR substrates, causing both a change in the shape of the curves and a translation in the position of the minimum. Note that, rather than display sensitivity as a function of complex index variations as two separate plots (one as function of the real part of the refractive index, the other as a function of the imaginary part), we chose instead to show maximal reflectance differences as a function of Methylene Blue concentration (Fig. 4c). As before, the slope of the curves yields the sensitivity. However, sensitivity is expressed here as a function of the experimental parameter (concentration), as opposed to an expression of instrument sensitivity alone – i.e. the result that is ultimately of interest in practice. Importantly for our application, LRSPR substrates clearly exhibit greater sensitivity to complex index variations (max dR/dC = 0.075 mg−1 for LRSPR, 0.038 mg−1 for SPR). To evaluate the sensitivity of LRSPR and SPR in experiments with living cells, we measured the response of HEK-293 cells to stimulations with different concentrations of lipopolysaccharides (LPS) from E. coli. Fig. 5a and b shows results from a typical experiment for each substrate type after 500 ␮g/ml LPS stimulation. For LRSPR (Fig. 5a), the response shows both a change in the shape of the SPR reflectance curve and a translation in the angular position of the minimum, indicating a complex refractive index variation in

V. Chabot et al. / Sensors and Actuators B 174 (2012) 94–101

99

Fig. 4. Typical experimentally measured angular scans for LRSPR substrates (a) and standard SPR substrates (b) at three Methylene Blue concentrations (0, 5 and 10 mg/ml); (c) average (n = 3) measured sensitivities (slope) at increasing Methylene Blue concentrations of 0 mg/ml to 10 mg/ml, in 1 mg/ml steps.

the dielectric. For SPR (Fig. 5b), only a change in the angular position of the minimum is observed, suggesting sensitivity to changes in the real part of the refractive index alone. Fig. 5c and d shows phase contrast images illustrating the morphological changes due to LPS stimulation. Here again, sensitivity in Fig. 5f is shown as a function of the experimental parameter (LPS concentration). As with the Methylene Blue model, these results demonstrate that LRSPR has greater sensitivity to complex index variations compared to standard SPR (slope of the logarithmic fit = 0.0186 for LRSPR, 0.0118 for SPR). As a negative control, both substrate types were washed with a detergent (Triton X-100 1% in water) at the end of each experiment to completely remove the cellular monolayer and return the substrates to a bare metal surface (Fig. 5e). As seen in Fig. 5a and b, the negative control surfaces exhibited an excellent response, demonstrating that the LPS-induced responses were not attributable to any degradation of the metallic substrates due to the presence of the cellular monolayer.

4. Discussion As observed in the experiments with purely real, purely imaginary, and complex refractive index changes, (Figs. 2c, 3c, and 4c, respectively), LRSPR exhibits higher bulk sensitivity compared to standard SPR. This arises from the greater proportion of the evanescent electric field located in the sensing dielectric medium for

LRSPR. These results also show that over the range of indices tested in the experiments, SPR sensitivity (slopes in Figs. 2c, 3c, and 4c) to both real and imaginary index component changes is relatively constant. This is not the case for LRSPR, where sensitivity to changes in the imaginary component is higher for small variations in the refractive index (Fig. 3c). As noted above, LRSPR measurements from cellular monolayers (Fig. 5a) show a profile that is typical of complex refractive index variations (translation, change in shape and minimum value), which implies that attenuation by scattering plays a significant role in the response. Scattering of light by cells results both from the micrometric dimensions of the cell as a whole (the cell is a scattering “particle” in itself) and from refractive index heterogeneities within the cell body, which range from 1.35 to 1.38 for the different cytoplasm organelles to as high as 1.40 for focal points [31–34]. Stimulation with LPS causes molecular and morphological changes within the cell that induce overall contraction and rounding of the cell body, as previously observed in phase contrast microscopy [14]. These changes in turn affect: (1) the nature of the scattering by the cells as a whole; (2) the profile of the complex refractive index within the cells; (3) the cell/fluid volume ratio in the medium. This last effect is particularly important: cell rounding and contraction increases the relative proportion of (optically homogeneous and transparent) medium within the range of the evanescent field, decreasing the overall scattering losses. As a result, the LRSPR response narrows and the minimum at resonance decreases under LPS stimulation (Fig. 5a), relative to the

100

V. Chabot et al. / Sensors and Actuators B 174 (2012) 94–101

Fig. 5. Typical experimentally measured angular scans for LRSPR (a) and SPR (b) substrates with an untreated HEK-293 cell monolayer, after 30 min LPS 500 ␮g/ml stimulation and after a Triton 1% wash; 20× phase contrast micrographs showing (c) untreated HEK-293 cells, (d) after 30 min LPS 500 ␮g/ml stimulation and (e) after Triton 1% wash, scale bars represent 20 ␮m; (f) average (n = 3–5) sensitivities (slope) after 30 min stimulation with increasing LPS concentrations on LRSPR and SPR substrates. The solid lines represent logarithmic fits with R2 values of 0.9928 for LRSPR and 0.9937 for SPR.

monolayer at rest. Conversely, SPR measurements (Fig. 5b) for the same cell type stimulation experiments show a profile that is typical of purely real refractive index variations (i.e. translation only). These results suggest overall that: (1) for LRSPR, cells must be modeled as complex index medium; (2) scattering by the cells is not significant within the range of the evanescent field in standard SPR.This correlates well with total internal reflection (TIR) microscopy where scattering is known to increase with the penetration depth of the evanescent field [25,35]. Cellular scattering of an evanescent wave has also been previously reported using sensing techniques with an increased probing depth, such as asymmetric waveguides and infrared surface plasmon spectroscopy [26,36,37]. Therefore, because of its greater penetration

depth and sensitivity to scattering, LRSPR is >50% more sensitive for detecting events that induce cellular morphology changes such as stimulation by a toxic agent, compared with standard SPR. Note that since the goal of this study was to compare the sensitivity of LRSPR and SPR, both substrate types were fabricated with near-identical metal film structures (32 nm Ag/5 nm Au, with the exception of the Ni adhesion layer). It is well-known, however, that optimal LRSPR substrates have a thinner metal film resulting in a lower effective index for the guided plasmon mode, further increasing the penetration depth and sensitivity [27]. Therefore, the sensitivity estimations reported here for LRSPR are a minimum only and optimized LRSPR substrates are expected to perform even better for cellular measurements. Moreover, due to the non-linear

V. Chabot et al. / Sensors and Actuators B 174 (2012) 94–101

LRSPR sensitivity to changes in the imaginary component of the refractive index, LRSPR is also expected to be particularly sensitive to the response of a cellular population achieving modest degree of confluence, such as neuronal cell lines or primary cell cultures. 5. Conclusions In this study, we showed that living cells must be modeled as a dielectric with a complex refractive index to take full advantage of LRSPR sensitivity to scattering losses within the penetration depth of the evanescent field. LRSPR is more sensitive than SPR for detecting events that induce cellular morphology changes. Therefore, LRSPR is particularly well suited for cell studies where the objective is to detect and quantify cell adhesion, the reorganization of the cell cytoskeleton, the redistribution of the cell organelles, or the variation in the density of adhesion sites the substrate. Acknowledgements The authors would like to acknowledge Dr. Pierre-Jean Zermatten for discussions about long range surface plasmons and Michael Lacerte for his help with the substrate fabrication protocol. This work was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC), the Canadian Institutes of Health Research (CIHR) and the Fonds Québécois de la Recherche sur la Nature et les Technologies (FQRNT). References [1] R.A. Yotter, L.A. Lee, D.M. Wilson, Sensor technologies for monitoring metabolic activity in single cells – Part I: optical methods, IEEE Sensors Journal 4 (2004) 395–411. [2] J. Homola, Surface plasmon resonance based sensors, in: O.S. Wolfbeis (Ed.), Springer Series on Chemical Sensors and Biosensors, Springer, Berlin, 2006, p. 251. [3] A. Shalabney, I. Abdulhalim, Sensitivity-enhancement methods for surface plasmon sensors, Laser & Photonics Reviews 5 (2011) 571–606. [4] X.D. Hoa, A.G. Kirk, M. Tabrizian, Towards integrated and sensitive surface plasmon resonance biosensors: a review of recent progress, Biosensors and Bioelectronics 23 (2007) 151–160. [5] X.-M. Zhu, P.-H. Lin, P. Ao, L.B. Sorensen, Surface treatments for surface plasmon resonance biosensors, Sensors and Actuators B: Chemical 84 (2002) 106–112. [6] R. Foldbjerg, D.A. Dang, H. Autrup, Cytotoxicity and genotoxicity of silver nanoparticles in the human lung cancer cell line, A549, Archives of Toxicology 85 (2011) 743–750. [7] B.H. Ong, X. Yuan, S.C. Tjin, J. Zhang, H.M. Ng, Optimised film thickness for maximum evanescent field enhancement of a bimetallic film surface plasmon resonance biosensor, Sensors and Actuators B: Chemical 114 (2006) 1028–1034. [8] C.F. Bohren, D. Huffman, Absorption and Scattering of Light by Small Particles (Wiley Science Paperback Series), Wiley-VCH, Weinheim, 1998. [9] H. Chew, D.S. Wang, M. Kerker, Elastic scattering of evanescent electromagnetic waves, Applied Optics 18 (1979) 2679–2687. [10] S. Ekgasit, A. Tangcharoenbumrungsuk, F. Yu, A. Baba, W. Knoll, Resonance shifts in SPR curves of nonabsorbing, weakly absorbing, and strongly absorbing dielectrics, Sensors and Actuators B: Chemical 105 (2005) 532–541. [11] H. Raether, Surface Plasmons on Smooth and Rough Surfaces and on Gratings, Springer-Verlag, Berlin, 1988. [12] M. Hide, T. Tsutsui, H. Sato, T. Nishimura, K. Morimoto, S. Yamamoto, et al., Real-time analysis of ligand-induced cell surface and intracellular reactions of living mast cells using a surface plasmon resonance-based biosensor, Analytical Biochemistry 302 (2002) 28–37. [13] Y. Yanase, H. Suzuki, T. Tsutsui, T. Hiragun, Y. Kameyoshi, M. Hide, The SPR signal in living cells reflects changes other than the area of adhesion and the formation of cell constructions, Biosensors and Bioelectronics 22 (2007) 1081–1086. [14] V. Chabot, C.M. Cuerrier, E. Escher, V. Aimez, M. Grandbois, P.G. Charette, Biosensing based on surface plasmon resonance and living cells, Biosensors and Bioelectronics 24 (2009) 1667–1673. [15] V. Yashunsky, S. Shimron, V. Lirtsman, A.M. Weiss, N. Melamed-Book, M. Golosovsky, et al., Real-time monitoring of transferrin-induced endocytic vesicle formation by mid-infrared surface plasmon resonance, Biophysical Journal 97 (2009) 1003–1012. [16] Y. Yanase, T. Hiragun, S. Kaneko, H.J. Gould, M.W. Greaves, M. Hide, Detection of refractive index changes in individual living cells by means of surface plasmon resonance imaging, Biosensors and Bioelectronics 26 (2010) 674–681.

101

[17] F. Liu, J. Zhang, Y. Deng, D. Wang, Y. Lu, X. Yu, Detection of EGFR on living human gastric cancer BGC823 cells using surface plasmon resonance phase sensing, Sensors and Actuators B: Chemical 153 (2011) 398–403. [18] S. Baumgarten, R. Robelek, Surface plasmon resonance (SPR) sensors for the rapid, sensitive detection of the cellular response to osmotic stress, Sensors and Actuators B: Chemical 156 (2011) 798–804. [19] C.M. Cuerrier, V. Chabot, S. Vigneux, V. Aimez, E. Escher, F. Gobeil, et al., Surface plasmon resonance monitoring of cell monolayer integrity: implication of signaling pathways involved in actin-driven morphological remodeling, Cellular and Molecular Bioengineering 1 (2008) 229–239. [20] P. Berini, Long-range surface plasmon polaritons, Advances in Optics and Photonics 1 (2009) 484–588. [21] P. Berini, Bulk and surface sensitivities of surface plasmon waveguides, New Journal of Physics 10 (2008) 105010. [22] M. Vala, S. Etheridge, J.A. Roach, J. Homola, Long-range surface plasmons for sensitive detection of bacterial analytes, Sensors and Actuators B: Chemical 139 (2009) 59–63. [23] C.-J. Huang, J. Dostalek, A. Sessitsch, W. Knoll, Long-range surface plasmonenhanced fluorescence spectroscopy biosensor for ultrasensitive detection of E. coli O157:H7, Analytical Chemistry 83 (2011) 674–677. [24] D. Ganic, X. Gan, M. Gu, Three-dimensional evanescent wave scattering by dielectric particles, Optik – International Journal for Light and Electron Optics 113 (2002) 135–141. [25] A. Rohrbach, Observing secretory granules with a multiangle evanescent wave microscope, Biophysical Journal 78 (2000) 2641–2654. [26] V. Yashunsky, V. Lirtsman, M. Golosovsky, D. Davidov, B. Aroeti, Real-time monitoring of epithelial cell–cell and cell–substrate interactions by infrared surface plasmon spectroscopy, Biophysical Journal 99 (2010) 4028–4036. [27] J. Dostálek, A. Kasry, W. Knoll, Long range surface plasmons for observation of biomolecular binding events at metallic surfaces, Plasmonics 2 (2007) 97–106. [28] E.D. Palik, Handbook of Optical Constants of Solids, Academic Press, San Diego, 1985. [29] D. Heumann, T. Roger, Initial responses to endotoxins and Gram-negative bacteria, Clinica Chimica Acta 323 (2002) 59–72. [30] A.O. Aliprantis, R.B. Yang, M.R. Mark, S. Suggett, B. Devaux, J.D. Radolf, et al., Cell activation and apoptosis by bacterial lipoproteins through toll-like receptor-2, Science 285 (1999) 736–739. [31] J. Bereiter-Hahn, C.H. Fox, B. Thorell, Quantitative reflection contrast microscopy of living cells, The Journal of Cell Biology 82 (1979) 767–779. [32] J. Beuthan, O. Minet, J. Helfmann, M. Herrig, G. Müller, The spatial variation of the refractive index in biological cells, Physics in Medicine and Biology 41 (1996) 369–382. [33] X.J. Liang, A.Q. Liu, C.S. Lim, T.C. Ayi, P.H. Yap, Determining refractive index of single living cell using an integrated microchip, Sensors and Actuators A: Physical 133 (2007) 349–354. [34] J.R. Mourant, M. Canpolat, C. Brocker, O. Esponda-Ramos, T.M. Johnson, A. Matanock, et al., Light scattering from cells: the contribution of the nucleus and the effects of proliferative status, Journal of Biomedical Optics 5 (2000) 131–137. [35] M. Oheim, Imaging transmitter release. II. A practical guide to evanescent-wave imaging, Lasers in Medical Science 16 (2001) 159–170. [36] R. Horvath, K. Cottier, H.C. Pedersen, J.J. Ramsden, Multidepth screening of living cells using optical waveguides, Biosensors & Bioelectronics 24 (2008) 799–804. [37] R. Horvath, H.C. Pedersen, N. Skivesen, D. Selmeczi, N.B. Larsen, Monitoring of living cell attachment and spreading using reverse symmetry waveguide sensing, Applied Physics Letters 86 (2005) 071101.

Biographies Vincent Chabot obtained his M.Sc. in 2008 in electrical engineering from the Université de Sherbrooke (Sherbrooke, Canada). He is currently a Ph.D. student in the Electrical and Computer Engineering Department at the Université de Sherbrooke, working on plasmonics and biosensors. Yannick Miron received his M.Sc. in 2008 in pharmacology from the Université de Sherbrooke (Sherbrooke, Canada). He works currently as a research assistant in the group of Michel Grandbois in the Pharmacology Department at the Université de Sherbrooke. His research interests are cellular applications for surface plasmon resonance. Michel Grandbois studied Biophysics at Université du Québec à Trois-Rivières (Trois-Rivières, Canada), where he obtained his Ph.D. in 1997. He is currently professor in the Pharmacology Department at the Université de Sherbrooke (Sherbrooke, Canada). He holds the Canada Research Chair in Nanopharmacology and Atomic Force Microscopy. His current research involves the use of atomic force microscopy and surface plasmon resonance to study cellular processes. Paul Charette obtained his Ph.D. in biomedical engineering in 1986 from McGill University (Montréal, Canada). He is currently associate professor in the Department of Electrical and Computer Engineering at the Université de Sherbrooke (Sherbrooke, Canada).